Between 30 and 50 million people in Bangladesh have been chronically exposed to high levels of arsenic, a class I human carcinogen, through contaminated drinking water. Although not clearly understood, based on current evidence, arsenic most likely exerts its carcinogenic effect through oxidative stress mediated DNA damage. While skin cancer and its precursor lesions are the most common arsenic-related conditions, there are wide variations in risk. We propose to examine the novel hypotheses that variants in the oxidative stress genes involved in arsenic-induced oxidative DNA damage and in DNA repair genes involved in the repair of this damage, are related to the risk of skin cancers and their precursor skin lesions, either independently or by modifying the effects of arsenic and these effects may be influenced by host antioxidant status measured by serum carotenoids, tocopherol and selenium. These hypotheses are supported by our strong preliminary data. Using existing resources (stored genomic DNA and comprehensive arsenic exposure and clinical data) from our ongoing prospective cohort study of 12,000 individuals, we propose to examine our hypotheses through a series of case-control and case-cohort analyses among 1,200 already identified skin lesions cases, 400 yet-to-be diagnosed skin cancer cases and 1,200 already identified random cohort members as controls. We will genotype the skin lesion, skin cancer and control subjects for 13 polymorphisms in the 4 oxidative stress genes (MPO, CAT, MnSOD, and GPX) and 6 DNA repair genes (OGG1, XPD, XRCC1, XRCC3, LIG1 and LIG4) using the high-throughput FP-TDI method established in our laboratory. The effect of the polymorphisms on arsenic-induced skin cancers and premalignant skin lesions, both independently and jointly with arsenic exposure, will be evaluated by estimating relatives risks from Cox's proportional hazards regression models. The proposed study is the first to investigate these novel hypotheses and has several scientific, methodological, logistic and practical strengths. Data generated will have implications not only for the massive public health issue relevant for millions of people in an underdeveloped country but also for arsenic exposed populations in other countries, including the US. [unreadable] [unreadable]